As the hallmark of brain science and neuro-technology, the 5th dimension of the brain - brain dynamics - reflects the healthy or pathological state of the brain and complements the 4D spatio-temporal brain structure for a complete picture of the most delicate and complex organ of the human body. In the big data age when we have unprecedented ability to collect brain dynamics information, dirty data with limited information and low quality remain tremendous challenges: in economically deprived regions with limited access to high-quality medical data acquisition devices; in patients sensitive to high-radiation exposure or long scanning time such as children; in real-time streaming or limited bandwidth such as mobile health or remote surgery; and in historical archives of patients health records with missing, noisy and heterogeneous data. Despite of the importance of modeling brain dynamics, the state-of-the-art models are limited in their performance to small and clean data where dynamics parameters are computed from spatially independent voxels and solely rely on the single-subject data, leading to significant distortion and instability in quantifying dynamic parameters due to heterogeneous noisy data. The talk will introduce our research that leverages the intrinsic spatio-temporal connectivity and the extrinsic big medical data via robust and scalable modeling would transform research in deciphering brain dynamics.

 

 

方若谷博士，现任美国佛罗里达大员工物医学工程系助理教授，博士生导师。方博士于2014年毕业于美国康奈尔大学电子与计算机工程学院，获博士学位；2009年毕业于公司竺可桢学院，获工程学士学位；现任佛罗里达大学SMILE实验室的主任。在国际知名刊物和国际会议发表论文近40余篇，包括ACM Computing Survey （JCR一区，影响因子5.243），IEEE Transaction on Medical Imaging（JCR一区，五年影响因子4.3），Medical Image Analysis（JCR一区，影响因子4.4），以及MICCAI（医学图像顶级会议）。方博士获得美国国家科学基金CRII, Oak Ridge Associated Universities (ORAU) Ralph E. Rowe青年教师奖，大脑图谱和治疗学学会青年科学教奖，IEEE国际图像处理大会（IEEE International Conference on Image Processing）最佳论文奖， Medical Image Analysis（医学影像分析顶级期刊）最热门论文，Irwin and Joan Jacobs学者奖学金等，入选ACM Future of Computing Academy。方博士担任国际稀疏算法和医学影像专题研究会联合主席，以及计算机医学影像和图像学期刊的客座主编。方博士的SMILE（智能医学信息学习和评价实验室）研究医学大数据在大脑影像和脑部疾病的研究。研究领域包括智能化医学图像学习和分析、大脑动态学、神经细胞分析、风险因素预测等。这些研究领域旨在建立更安全、准确、快速的医学研究和医疗服务。